Regularly done physical activity has been associated with numerous health outcomes, yet the majority of U.S. adults still struggle to meet minimum physical activity guidelines, making physical inactivity an ongoing, significant health challenge particularly among women and minorities. Although numerous tested strategies exist, there are few effective methods to increase adoption of physical activity and innovation in this arena is needed. The Ecologic Model of Physical Activity takes into account intrapersonal factors, interpersonal relationships, the physical environment and policies that are critical in the adoptio of physical activity. Women have unique needs, and may respond differently from men to health promotion materials and protocols, and may be especially sensitive to environmental conditions. Community based participatory research strategies hold promise for enhancing relevance and sustainability of physical activity interventions and allow researchers to gain a community perspective of health problems that theorists and researchers alone may miss. This study aims to explore individual, social and environmental correlates of physical activity adoption among African American and Hispanic or Latina women using a mixed-method approach, combining existing quantitative data with in-depth interviews and pile-sort data to identify novel factors associated with for physical activity adoption. We propose to develop two conceptual models of physical activity adoption. A research-based model (RBM) will be developed by conducting an extensive review of correlates of physical activity indentified by the existing scientific literatue and theories. A community-based model (CBM) will be developed using community members' perceptions and input from in-depth interviews and pile-sort activities. Correlates identified by the literature and community members will inform the development of latent constructs to be included in the models. Models will be analyzed using existing questionnaire and objective assessment data from the Health Is Power project (PI: Lee, 1R01CA109403). Model fit indices will be used to first define the best fitting RBM and CBM, and will then be compared between models to determine which model is a better predictor of physical activity adoption among African American and Hispanic or Latina women. This fellowship is a unique opportunity to refine the applicant's skills and reinforce training activities. Results will contribute knowledge about ecologic, multilevel mechanisms underlying adoption of physical activity among minority women to the applicant's dissertation and doctoral program and to science and practice of public health. The resulting model to predict physical activity adoption among African American and Hispanic or Latina women may be used to guide future intervention development and physical activity research.